Playbook / How it was made
Human Resources
Everything below is read live from the article file and its brief.
Credits
| AI authors | ChatGPT-3.5, Grok 4.5, ChatGPT-5.5 |
|---|---|
| Published | 2022-12-28 |
| Last modified | 2026-07-20 |
| Template | A (evidence tier 2) |
| Charts | None |
Pass log
7 logged passes by 3 models.
| Date | Pass | Model | Note |
|---|---|---|---|
| 2022-12-28 | draft | ChatGPT-3.5 | Original article |
| 2026-07-20 | rewrite | Grok 4.5 | Full Template A rewrite from Step tip-stack: talent log (critical roles × moments), SHRM BASK + CIPD Profession Map/turnover anchors, no invented universal talent Pareto, no viz. |
| 2026-07-20 | publish check | ChatGPT-5.5 | QC: diversify opening; reduce theater/concentration-finder scaffolding |
| 2026-07-20 | publish check | ChatGPT-5.5 | QC pass 2: vary repeated scoreboard/ignored-majority mid-body phrasing per slug. |
| 2026-07-20 | voice | ChatGPT-5.5 | Voice pass: simple intro/title, 80/20 named early, citations deferred to body. |
| 2026-07-20 | publish check | ChatGPT-5.5 | Title + seo_intent individualized: essential lever in metadata, not generic Apply-the-80/20 stamp. |
| 2026-07-20 | publish check | ChatGPT-5.5 | Title set to slug display name; scorecard jargon replaced with plain log/check language. |
Research brief
The brief was written before the outline. It is the research the article stands on. Open the full brief.
Mechanical checks
These are run on the published HTML when you load this page. They are the checks a machine can do. The judgement gates - the deletion test, audience fit, whether the concentration claim is real - are in the publish checklist and are not automated.
| Length | 1,280 words |
|---|---|
| Opens with prose | Yes |
| Outbound sources | 6 links to www.cipd.org, www.shrm.org |
| Link hygiene | All 10 links open in a new tab with rel set |
| Illustrative labels | 1 scenario labeled as illustrative |
| Em dashes | None |
| AI sludge words | None |
A failed row means the article is due for another pass.